This commentary was written by Himadri Pandey, is an applied scientist for Lucas Systems. The views expressed here are solely those of the author and do not necessarily represent the views of Modern Shipper or its affiliates.
Done right, good slotting decisions can reduce labor costs (picking and replenishment efficiency); improve worker safety and ergonomics; mitigate product damage; and enhance warehouse space/capacity utilization.
‘Optimized slotting’ is the ideal
Warehouse slotting is the activity of determining the best storage location for each product in a warehouse. While the concept of slotting and its related benefits are generally easy to grasp, doing it well and consistently enhancing your slotting can be a significant challenge. Why? Because it can involve literally thousands of products and locations, numerous other input factors and multiple objectives that contribute to making it a multi-layered, incredibly complex problem.
Slotting is typically done by re-organizing the entire warehouse a couple of times a year and moving products in ways that place them in more optimal locations. This is considered a ‘static’ approach to slotting. The challenge with this approach, of course, is that things change every day.
To give DCs the ability to react to changes by re-slotting items is the process of dynamic slotting. This advanced approach is a new way to think about how best to optimize product locations given all the variables.
More recently, the use of data and AI have shown promise in helping to make slotting processes more dynamic so that pick locations are assigned with the ultimate efficiency.
Get more dynamic with your slotting
Let’s look at three situations when implementing a more dynamic slotting process utilizing AI, which may be worth exploring for your DC.
- Rapidly changing SKU velocity. Slotting is highly reliant on the velocity of SKUs, which changes with time, season, etc. If your warehouse has products in which the SKU velocity undergoes rapid change on a weekly or daily basis, then it would be an ideal fit for dynamic slotting. Dynamic slotting can handle fast moving products by pushing them into more easily accessible spots. In dynamic slotting, when the slotting moves are made every week, one can take advantage of the updated velocities, as well as the SKU demand forecast. In warehouses where the demand forecast is available or can be predicted, such as retail and groceries, this information can be utilized in the dynamic slotting process. If you compare that with more traditional static slotting, one can take advantage of this available information only a few times a year. Depending on the urgency or time frame of the forecast data, it may be outdated by the time the next slotting update is scheduled.
- Disruptions due to high volume movement. Moving products a couple times a year with static slotting can disrupt the flow of the warehouse for weeks. Typically, specific time chunks will be allotted in the facility for the complete re-warehousing to take place. When such a high volume of products are being moved in the warehouse, it is nearly impossible to continue the picking process. This disruption can lead to a substantial loss of precious time and resources. By contrast, with dynamic slotting, one can make a few changes every day, either during the workday or at the end of the day. If workers have a few moments, they can make slotting changes (moving 5-10 SKUs) and proceed to continue their job. Moving 5-10 SKUs every day in a 4,000 SKU capacity warehouse will not interrupt the daily workflow. Moreover, during a period of a few weeks, all the low hanging fruit (high impact SKUs) can be captured.
- Ability to realize cost reductions from moving only high impact SKUs. Dynamic slotting offers a lower cost of re-slot to static slotting (which is naturally higher because all the products are being moved). With dynamic slotting, only products which have high impact are moved. This results in fewer products being moved, leading to reduced costs of re-slotting. A dynamic slotting tool can help to determine the highest cost savings factor, whether it comes from picking travel cost or put-away travel cost.
Dynamic slotting’s benefits and use cases
Static slotting benefits are realized during the first few weeks of implementation but then start to diminish over time because of ever-changing SKU velocities. On the other hand, dynamic slotting benefits are seen more frequently and consistently realized because products are continuously being moved based on the latest velocities.
Dynamic slotting might eliminate the need for expensive engineering studies or training and expertise to use complex software. Our Lucas system, for example, applies advanced machine learning algorithms to SKU velocity, SKU affinity, product/slot information, pick paths and other data to recommend the best locations for your inventory. It does not require a detailed CAD drawing of the warehouse, as is generally necessary with conventional slotting software.
What are some of the potential benefits you might see through the use of dynamic slotting? In most cases, AI-based slotting can deliver 10-20% labor cost savings, a 1-5% improvement in accuracy and 20-40% increase in throughput.
If your warehouse meets any of the below factors, dynamic slotting is worth investigating:
- Your SKU velocities are constantly changing
- You are in a market affected by seasonality
- Your business provides forecasted data on SKUs
- You have a small warehouse and are constantly moving products from put-away to pick faces
- You cannot afford to interrupt the workflow in the warehouse
As warehouse operators continue to transform and seek competitive advantages, frequently evaluating and implementing new processes and innovations will be key to continued profitability and sustainability. Utilizing dynamic slotting in your organization could help you increase productivity, improve warehouse space utilization and reduce picking and replenishment labor costs.
About the author
Himadri Pandey is an Applied Scientist for Lucas Systems where she develops algorithms, builds predictive analytics, and designs novel solutions for warehouse workflow optimization. Prior to joining the supply chain industry, she worked on high energy particle physics and spacecraft engineering problems. She has a B.S. from the University of Cincinnati with a major in Computer Science and minors in Physics and Math. Passionate about machine learning, Himadri is starting on her Ph.D. at Georgia Tech.